[gpalmer]

I think I said that at the top of the page, just in English!

 

[taoster]

Quote:

Originally posted by gpalmer I think I said that at the top of the page, just in English!

yeah i know! LOL

just backing you up.

you didnt think i got to 1000 posts by writing insightful posts did ya?

 

[sTaTIx]

Quote:

Originally posted by Mr.PD ********. I have satellite TV. I have seen digital artifacts on the tv, little blocks of color, big blocks of color and I have seen it get so bad as to be unwatchable. But the signal is still trying to come through. This was not caused by the analog equipment. It was caused by snow on the dish, and/or snow in the air. Many of the times this has happened just before losing the signal to the point of no picture on the tv, but sometimes it will clear up after a few moments. The signal is not as on/off as you seem to imply.

I guess this is the digital video version of the "jitter" that we sometimes hear when there are errors on the digital side of the equipment. There is snow on your dish which is causing some bits of the data to not come through. Just like when you apply shock to a portable CD player, it starts skipping because it didn't read portions of the song data.

In either case, you'll see that it's not hard to see when a digital device is misfunctioning, at all. If video is not being properly transmitted, you'll see the aforementioned "blocks of color" (pixels?), and with audio, you'll hear stops and skips in the music.

 

[JaZZ]

No – missing data or a poor error correction aren't the equivalent of jitter. Jitter is an uneven data flow, out of rhythm. All data are preserved, but the analog curve resulting from the loudness values transported by each of the 44,100 samples/sec is slightly twisted related to the curve which is actually meant by the recording – because the sample points are slightly shifted in time. Thus the frequency spectrum of the curve is falsified.

Missing or corrupt data and their correction algorithms are another, maybe secondary factor in the domain of digital colorations.

 

[taoster]

Quote:

Originally posted by JaZZ No – missing data or a poor error correction aren't the equivalent of jitter. Jitter is an uneven data flow, out of rhythm. All data are preserved, but the analog curve resulting from the loudness values transported by each of the 44,100 samples/sec is slightly twisted related to the curve which is actually meant by the recording – because the sample points are slightly shifted in time. Thus the frequency spectrum of the curve is falsified. Missing or corrupt data and their correction algorithms are another, maybe secondary factor in the domain of digital colorations.

you are right JaZZ.. how could i get it all so mixed up

so sound artifacts are caused more by the difference in digital "packet spacing" rather than the actual data content.

what i dont understand is, why is it so difficult to remove jitter all together? surely by now, they're able to transport 44.1khz reliably (by some other way)?

edit: thinking.. surely if both the transport and DAC itself uses a hard-disk based buffer system and the data are transported as data and not synchronous wave, jitter can be removed entirely! Considering the cost of high audio equipment, how hard can it be to do something like that??

 

[JaZZ]

I remember a test in the German «Stereoplay» years ago. Between two obviously different sounding transports not a single bit of difference was found in the data sum check, but a significant difference in the jitter spectra. The conclusion was that the contribution of errors and their correction to sound alterations be negligible, in contrast to jitter, which was highly suspected to be responsible for them.

I'm not an expert in digital, I have just basic knowledge (so I can't comment your problem solving proposal). BTW, I've always suspected that the CD format isn't accurate enough for sensitive ears, but never considered sound variations in the digital domain possible.

 

[gpalmer]

Jitter reduction/prevention can be acheived either by a more accurate clock on the player, such as Audiocom's SuperClock or by incorporating a buffer with the receiving DAC. Many third-party DACs do this now and all they do is store the little packets of data they receive and reclock them so they are fed into the DAC at the correct intervals.